This is a fake news classification project, using TFIDF and pre-trained w2v embedding as separate sets of features, along with text sentiment scores, to classify news text as fake or real.